Abstract The growth and retraction of dendritic spines with synapse formation and elimination is thought to underlie experience-dependent changes in brain circuitry during development and in the adult brain, and also may play a role in neurodevelopmental disorders. While much focus has been given to spine growth and associated synapse formation as a key step in the development of brain circuits, the mechanisms of spine retraction and synapse disassembly are ill-defined. During late postnatal development, after initial connectivity has been established, dendritic spine densities decrease and half of all synapses are lost in some regions of the cortex. This period coincides with a period of intense learning, suggesting that spine and synapse elimination may have an integral role in learning and memory. Indeed, many neurological disorders that result in mental retardation have been associated with spine loss and spine morphology changes. The primary goal of the proposed studies is to determine the mechanisms that govern the retraction of dendritic spines and the disassembly of synapses during brain development, plasticity, and disease. We have recently shown that spine shrinkage is initiated by both input-specific and locally competitive activity patterns that lead to synaptic weakening and that spine retraction is tightly coupled to disassembly of the postsynaptic density. We currently have three aims. First, we will determine the molecular mechanisms that drive input-specific spine shrinkage and retraction and synaptic weakening. Second, we will define the cellular and molecular signaling mechanisms by which competition between neighboring synapses drives spine shrinkage and retraction, and what role these competitive mechanisms play during circuit plasticity in vivo. Finally, we will determine how synaptic interactions and molecular composition are regulated during input-specific and heterosynaptic spine shrinkage and retraction. To achieve these goals, we will use focal photolysis of caged glutamate to stimulate individual spines, combined with electrophysiology to measure spine synapse function, calcium and fluorescence lifetime imaging to measure real-time signaling in dendritic microdomains, and electron microscopy to monitor pre- and postsynaptic ultrastructural changes and molecular alterations during spine retraction. The combined use of advanced two-photon imaging techniques, electrophysiology, and electron microscopy at single synapses will provide an innovative and powerful way to identify the mechanisms that govern the retraction and disassembly of spine synapses. Results from our experiments will rigorously address the mechanisms of spine retraction and synapse disassembly, thereby filling major gaps in our current understanding of neural circuit refinement during development and experience-dependent plasticity. Ultimately, basic knowledge of the mechanisms of spine elimination has strong potential to facilitate the development of therapeutics for neurological diseases.